Hi, i have an idea.
Could you write a text file next to the enhanced video file with all the parameters used for the review/enhanced video. For Example, the result would be Video.mp4->Video_enhanced01.mkv+Video_enhanced01.mkv.txt
Where .txt file would contain info like:
Proteus v3
Proteus parameter 1 = 23
Proteus parameter 2 = 35, etc.
Video encoder H265, Q=15, audio encoder AAC, Q=128bits/s.
Will AVS reads no longer be supported?
Some results are better when filtered by Avisynth and then processed by VEAI, so I would like to see continued support.
Also, previous versions process in sRGB (8bit), so even when processing 10bit or higher data, it was still 8bit internally, have those been improved?
High bit processing is useful to prevent banding.
Japanimation is often produced at low resolutions such as 720p and scaled to 1080p.
For such works, it may be better to scale them down to the production resolution and then upscale them with VEAI to obtain a better result.
If AVS is available, the following workflow is possible.
Avisynth files have never been supported in Video Enhance AI. Can you please confirm if you were able to open some files in Video Enhance AI but you can’t open them in this alpha? If so, can you share with me a video or two? Thanks!
How do I transfer data from ffmpeg veai to avisynth?
something like this:
ffmpeg -i “video.mp4” -filter_complex veai_up=… -f yuv4mpegpipe - | ffmpeg -i script.avs …
which line for pipe input in script.avs?
Using deinterlacing method ffmpeg gives very weird results on upscaled video. As you can see there are lot of jittery/blocky artifacts in the video. I was wondering when we are going to get this fixed, what things can be done about this, and when we are getting some model updates? Many thanks for the new update btw!
